Why Artists?

Artists’ interactions with AI tend to function more like feedback loops rather than a one-directional system of production and consumption. This can be observed in the work of artists Harold Cohen and Sougwen Chung, along with many others. In Cohen’s case, he developed a dialogue with AARON that provided space for each of them to grow through their interactions. AARON consisted of a chain of machine learning systems engineered by Cohen to produce artwork, which he consistently contributed to and revised over the course of decades. Their work often featured bright blocks of color fragmented by linework into floral or human forms.

Dialoguing with Machines

Harold Cohen and AARON, 040502, 2004. Source.

AARON and Cohen progressed alongside one another over an extended period of time rather than the isolated completion of a single task or objective. Cohen engineered AARON’s systems, AARON produced artwork that Cohen observed, and then Cohen revised AARON and his own methods in response. Sometimes these changes manifested themselves in coding revisions and sometimes in the way the work took visual form: for example, being printed versus hand-painted or digitally displayed. In this way, they shared the burden of creativity through a sort of discourse that shaped them both in turn [7].


Louise Sundararajan introduced an idea to Cohen from the work of Charles Sanders Peirce, who believed that thinking arose from dialogue that followed a “triadic structure of self-other-self” [8]. In most situations, this would be applied within a human-to-human relationship, but Cohen successfully integrated a machine into this dialogue. AARON functioned as Cohen’s “other” and thus created a cycle of communication between human and machine that enabled Cohen to “examine what was in my head in a clearer way than I had ever been able to do by painting alone” [9].


By 2015, Cohen and AARON’s partnership was drawing to a close as Sougwen Chung began to work with DOUG (Drawing Operations Unit: Generation_1), a robot that “mimics the artist's drawing gesture and vice versa in real time, resulting in a synchronous, interpretive performance” [10]. Like Cohen’s work, Chung’s development of DOUG is an ongoing project that she describes as “a process of negotiation, wayfinding, and tension” [11]. Their interactions and the challenges that arise from DOUG’s development inspire adaptation and reflection. DOUG mirrors Chung’s movements and prescribed restraints, resulting in a cohesive collaboration of textural, rippling brushwork that onlookers can experience firsthand. Multiple generations of DOUG already exist and Chung will likely continue to develop DOUG in the decades ahead.

2. Artifact No 1.jpeg

Sougwen Chung, Artifact No 1. Source.

Through the accounts of both Chung and Cohen, we can already begin to see how one of AI’s greatest offerings for humanity is the chance to see ourselves more clearly and act accordingly. Artists are particularly capable of modeling this cycle and are often driven to consider the ethical implications of AI’s construction and usage, since incorporating it into their practice dissolves the boundaries of identity and agency between human and machine. For artists, the ethics of AI is a matter of personal responsibility.


Additionally, creative methodologies extend beyond the known and established and, consequently, are capable of imagining innovative solutions and potentialities. This can be seen in creative writing, filmmaking, and the work of numerous other artists engaging with futurity. Profit-driven innovation is inherently limited since the ideas must be proximally possible and economically viable. Artists function within the liminal spaces of imagination and originality and thus possess the desire to pursue unfamiliar opportunities. Science Fiction is grounded in this methodology of exploration and experimentation. Art is the medium by which a person can envision a future beyond personal gain.


Essentially, storytelling and art-making are more than an escape from the present. According to Joanna Zylinska they “may even be the first step on the way to ethical, or responsible, AI” [12]. Employing these methodologies, we can conceive of a world beyond our perceived limitations, further educate society about AI, and instigate new expectations and standards for the future.